23M043 AI identification of light pollution sources

Light pollution is a global environmental issue in which excessive artificial outdoor lightings affect the natural environment and the ecosystem. Among different forms of light pollution, light trespass needs immediate attention due to its potential adverse impacts on human health. The problem of light trespass is particularly evident in a densely-populated metropolis like Hong Kong. Using Artificial Intelligence (AI), computation methods are developed to detect, classify and quantify the light sources in images and videos taken at the street level. Results can be used as a light condition analysis for analyzing whether light emitters are sourced from commercial buildings or street areas to determine the effects of light pollution on residents’ views. As a citizen science project, members of the public are encouraged to contribute images or videos to expend the coverage of the study. The lecture introduces light pollution, citizen science projects and the research conducted by the University of Hong Kong. Supported by the RAE 2026 Fund of The University of Hong Kong.